Workshop: How to choose a research topic

How do you find an original research topic and create impact, when there’s a deluge of information or lots of researchers in your field?

In this workshop, we look at some approaches that you can put into practice to transform the way that you review research papers, to quickly identify the gaps. Challenge your thinking, gain inspiration and find your research opportunities, even if the problem has been addressed in different ways before.

You will need:

• A problem or research question in mind
• A collection of unread research papers, related to your research problem or question

Part 1: Critically examine the research question

Research gaps can be identified quickly, just by looking at the research goal (the question or problem being addressed) and the assumptions
made by the author. From this information you will be able to see the context, limitations of the research and help clarify your own goal.

Exercise: What’s the problem?

Research opportunities start with a problem! So, in this exercise we’re going to explore the problem and how you would go about solving it, before reading anything!

1. Think about the problem and write down a list of all the aspects that need to be addressed to solve it
2. Select a paper and find the section that defines and details the problem. Tick off how many of the aspects you listed match and add the ones you missed to your list
3. STOP reading

ASK YOURSELF – do the aspects you missed actually matter?

Useful tip – start with the oldest paper so you can see how the problem and assumptions evolve over time!

Exercise: What assumptions were made?

Staying with the same paper, it’s now time to review the assumptions and limitations of the research.

1. Identify and note down the assumptions made
2. Assess and highlight which of the assumptions hold true

ASK YOURSELF – could you tackle the problem if you take out some of the assumptions?

Discovering research is based on assumptions that could be challenged or removed can create a new research opportunity. To find out what happens if it was done differently and take it in a new direction.

1. Now look at the way the assumptions have been stated and categorise them into what you would consider to be ‘fuzzy’ vs succinctly formulated (using mathematical or discipline-specific formalization)

ASK YOURSELF – what could you do to make the ‘fuzzy’ assumptions complete and would this bring new insight or add another dimension to the problem?

Once you have reviewed your first paper, systematically go through each of your papers, just reviewing the problem and assumptions. Start from the oldest to most recent to see changes over time and compare your findings.

DO NOT READ ALL OF THE PAPER

Summary: Get inspiration and build upon the limitations of prior research to bring a new perspective to a problem. Focus on the problem and assumptions to identify the gaps to establish whether you are going to build upon existing research or start from scratch.

Part 2: Critically examine the approach

Avoid the no-significant-contribution trap by thinking like a reviewer. Asking these questions will help you to examine the approach more thoroughly and fuel your own ideas:

1. Why does the gap exist? Is additional proof, evidence or better measurement required?
2. Was the research incomplete or not filling the gap?
3. If it was reproduced, am I confident that I would get the same results?

Exercise: What are the key hurdles to overcome?

The aim of this exercise is to surface past constraints and new advancements to understand the hurdles and barriers that others have experienced when trying to solve this problem.

1. Find approaches that claim to be novel or original
2. Take note of the hurdles described and the nature of them i.e. are they mathematical, computational, experimental or evidential

ASK YOURSELF – how original is it and does it fully address the problem?

1. Examine the methods chosen by the author
2. Assess each approach based on correctness, completeness, & reproducibility

ASK YOURSELF – is there any logical fallacy, inconsistencies or lack of supporting evidence?

1. Think and discuss whether all the steps in the method could be reproduced and deliver the same results?

Once you have reviewed the approach of all relevant papers, collate your findings to review common hurdles and identify what is difficult to overcome and why.

DO NOT READ ALL OF THE PAPER

Summary: Deeply understand the challenges faced in prior research to identify new ways to overcome past barriers or to complete the gaps.

 

Part 3: Critically examine the evaluation

And finally, what impact did the research have? It’s time to critique the claims and review the evaluation of the results. The goal is to assess:

1. Is the approach showing the expected behaviour?
2. Can we find any new insight(s) by analyzing the results?
3. Is the problem (hypothesis) being supported by the results?

Exercise: Are there gaps in the evaluation setup?

In this exercise we are going to be focusing on the data points and metrics, looking at what’s missing and considering what evidence really matters.

1. Locate the results and the datasets that have been used by each paper
2. Identify whether standard (publicly available) datasets have been used or whether the author has created their own
3. Review what’s missing or if there could be any bias

ASK YOURSELF – Is it a good quality dataset to evaluate the approach or could a better dataset be created?

1. Identify whether existing evaluation parameters and metrics have been used

ASK YOURSELF – Can you find any loopholes and prove these metrics of these parameters are not worthwhile?

1. Look out for new questions, where there are no standard parameters or where there is disagreement as to which is the best approach

ASK YOURSELF – Is there a better way to measure a parameter and why is this needed?

Systematically review the results of each paper, collate your findings and compare what the results are based on to identify differences, controversies, opportunities and gaps.

Summary: Evaluating the results provides additional insights and opportunities to identify gaps and new opportunities to contribute to the research community. Whether that be new datasets, parameters or through finding loopholes that require problems to be addressed in different ways.

Conclusion:

Identifying the gaps requires investigation and systematic examination of different aspects of research papers. Reviewing 1 section of a paper at a time, capturing the insights. After each section, reviewing, filtering and comparing your findings. Looking for loopholes, questioning the claims and seeing how it compares to how you would approach the problem.

Watch the full series: